= Solution
\b[A physiological role for an ER–mitochondrial contact requires more than demonstrating close membranes: changing the contact must alter a measured transfer or cellular process.] The most persuasive evidence combines quantitative imaging, selective perturbation, functional measurements and rescue. Physical proximity, tethering, <lipid> transfer and <calcium> signaling are related but experimentally separable claims.
A <membrane contact site> brings distinct organelles close without fusing their membranes. At <endoplasmic reticulum-mitochondrial contact sites>, the <endoplasmic reticulum> and <outer mitochondrial membrane> remain separate, preserving their different lumenal spaces and <lipid> compositions. <electron microscopy> can establish apposition and measure distance or contact extent, but a static image cannot establish exchange. Fluorescent colocalization at ordinary optical resolution is weaker still: overlap can occur at distances larger than those required for a molecular bridge. Report the contact definition and consider organelle size, distribution and shape when comparing contact frequency.
<calcium> transfer provides functional evidence. Targeted calcium-sensitive reporters showed that the mitochondrial surface can experience a much larger local signal after ER <calcium> release than a bulk cytosolic reporter detects.<Close contacts with the endoplasmic reticulum as determinants of mitochondrial Ca2+ responses by Rosario Rizzuto and others>[pp. 1763–1766, organelle-targeted calcium reporters and local responses after ER release]{ref} This explains how local transfer can occur despite a modest spatially averaged cytosolic increase. The <inositol trisphosphate receptor> releases <calcium> from the ER, and the resulting local signal can drive mitochondrial uptake. Control ER <calcium> loading, stimulus strength, reporter calibration and mitochondrial <membrane potential>, since reduced mitochondrial uptake can result from a weak driving potential rather than increased interorganelle separation.
A mammalian tethering experiment connected architecture to neuronal signaling. Loss of <PDZD8> reduced ER–mitochondrial contacts and altered mitochondrial <calcium> uptake and cytoplasmic <calcium> dynamics after physiological stimulation in neurons.<ER-mitochondria tethering by PDZD8 regulates Ca2+ dynamics in mammalian neurons by Yusuke Hirabayashi and others>[pp. 623–630, contact measurements, PDZD8 perturbation and neuronal calcium responses]{ref} Rescue and measurements of other organelle properties help distinguish a tethering effect from general cellular disruption. The important inference is that contact architecture contributes to <calcium> handling in the tested cells. It does not establish that all mammalian contacts use <PDZD8>, or that every effect of the <protein> is explained exclusively by one calcium-transfer route.
A complementary genetic strategy identified the yeast <ERMES> complex using a synthetic ER–mitochondrial tether. A designed bridge could suppress defects of mutants affecting the native contact machinery; localization and genetic interaction analysis connected the identified components to the two organelles.<An ER-mitochondria tethering complex revealed by a synthetic biology screen by Benoît Kornmann and others>[pp. 477–481, synthetic-tether rescue, screen and contact-complex localization]{ref} Rescue by a physical bridge supports a contribution from apposition itself, but incomplete rescue can mean that the native complex has additional functions or that the artificial bridge has an inappropriate geometry. Conversely, rescue of growth alone does not prove that the native complex is itself a <lipid> channel. Several downstream defects can improve when two organelles are brought together.
<lipid> biosynthesis offers a biochemical transfer assay. <phosphatidylserine> made at the ER can be converted into <phosphatidylethanolamine> by a mitochondrial <enzyme>. Label incorporation into the precursor and product can therefore report interorganelle flux. Experiments with isolated membrane preparations found reduced transfer-dependent conversion when <ERMES> function was impaired.<A phospholipid transfer function of ER-mitochondria encounter structure revealed in vitro by Rieko Kojima and others>[main results and in vitro phospholipid-transport assays, labeled PS-to-PE conversion]{ref} This is stronger than measuring steady-state <lipid> amounts, because a pool can remain constant while synthesis and turnover change together. Nevertheless, conversion is a compound readout: synthesis of the labeled precursor, delivery to the mitochondrial <enzyme> and the <enzyme>'s own activity all matter. Measure those steps separately and normalize to organelle amount.
The interpretation of such an assay can be expressed as a simple mass balance. For labeled mitochondrial product $P^*$,
$$
\frac{dP^*}{dt}=v_{\mathrm{conversion}}^*-v_{\mathrm{loss}}^*.
$$
If loss is negligible over an early interval and conversion is rapid once precursor arrives, the initial product slope approximates transfer flux. If conversion is slow or the product is rapidly exported, that approximation fails. This is an assay model, not a claim that all <ERMES> experiments meet the limiting assumptions. Direct transfer assays with purified candidate components and distinct donor and acceptor membranes can test molecular sufficiency, but must exclude membrane fusion, leakage and <lipid> exchange unrelated to the candidate.
Contacts are also associated with mitochondrial division. Live imaging and ultrastructural analysis showed ER tubules marking mitochondrial constriction and division sites, with contact occurring before completion of fission.<ER tubules mark sites of mitochondrial division by Jonathan R. Friedman and others>[pp. 358–362, time-resolved ER association with constriction and division]{ref} Temporal ordering supports an upstream role but does not by itself identify the force generator. Separate loss of contact from loss of the fission machinery, quantify completed divisions rather than constrictions alone, and control mitochondrial network geometry. A global increase in fragmented mitochondria can reflect increased division, decreased fusion or impaired organelle movement, so it is an ambiguous endpoint without a dynamic assay.
The physiological interpretation should also account for redundancy. Yeast <ERMES> and mammalian tethers are not interchangeable molecular labels, and compensating contact routes can mask loss-of-function phenotypes. Acute depletion can reduce developmental compensation; a separation-of-function mutant can distinguish tethering from another activity. Artificial tethers should be titrated, since excessive or overly tight contact can create abnormal transfer or constrain organelle motion. A rescue that restores physiological spacing and function is more informative than simply maximizing contact area.
<calcium> handling, <lipid> supply and division can interact. A primary loss of <lipid> transfer might change mitochondrial shape and respiration, secondarily affecting <calcium> uptake. A primary <calcium> defect might change metabolism and later <lipid> pools. Short time courses, direct functional assays and selective rescue help establish causal order. Compare contact morphology, <lipid> flux, <calcium> dynamics, respiratory competence and viability rather than treating them as interchangeable readouts. \b[The combined evidence supports contacts as regulated sites of material exchange and organelle organization; the experimental task is to assign a particular function to a particular contact mechanism without confusing proximity with transport or a secondary mitochondrial defect.]
Back to article page
